Project Summary ? Project 1 Melanoma cells adapt well when transplanted to immune-deficient mice and the group has established >450 PDX, which allow us to identify in-depth those subgroups of tumors that are most suitable for selected drugs. The PDX have been extensively characterized for their growth properties in vitro and in vivo including metastasis formation, somatic mutations in 108 genes through targeted sequencing, and expression of 305 proteins related to signal transduction. The models are superbly suited to investigate two major pathways that are consistently activated in melanoma and in most other cancers: MAPK and PI3K pathways. Our overall working hypothesis is that we can predict therapy responses based on the genetic, genomic, and biologic signatures of tumors. We expect that we can define, short of a complete personalized approach, subgroups of melanomas that will respond to new therapies and new combinations of therapies in a predictable way. Using drugs that are available through NCI-CTEP, we will define in the first aim those melanomas that are most sensitive to inhibitors of MET in combination with MAPK inhibitors. We will also investigate multi-kinase inhibitors as single agents and in combination with MAPK inhibitors. Our working hypothesis is that in subgroups of melanomas, MET as well as other receptor tyrosine kinases (PDGFR, VEGFR, c-KIT, FGFR1) and non-receptor kinases (SRC) are important targets. We expect to develop biomarkers that will pinpoint those melanomas that are most (or least) susceptible to kinase inhibition. In the second aim, we will define melanomas that are most sensitive to combined inhibition of the MAPK and PI3K pathways. Our working hypothesis is that a subgroup of BRAF-mutant melanomas exists, which is highly susceptible to the combination of BRAF/MEK with PI3K inhibitors. Based on our preliminary studies we will use loss of PTEN as one biomarker for selection but expect further refinements, especially for tumors resistant to BRAF and BRAF/MEK inhibitors. Since IGFR-1 signals predominantly through the PI3K pathway and is in selected melanomas a key player in acquired resistance, we will include the appropriate CTEP inhibitor(s). At the same time, we will determine whether isoform specific PI3K inhibitors are more appropriate in combination with MAPK inhibitors. We expect to define biomarkers for the selection of both treatment-nave and BRAF/MEK inhibitor-resistant tumors. Our studies should provide a strong rationale for initiating clinical trials.